The overarching goal of this research is to improve the understanding of factors explaining one of the most pervasive results of social epidemiology: the migrant mortality advantage, i.e., the observation that migrants tend to exhibit lower mortality than the non-migrant population of their host country. In spite of the extensive literatue on the topic, this mortality advantage remains questionable and poorly understood, in part because of the very nature of the migrant population: a population that is highly mobile and difficult to capture correctly in data sources. Studies documenting the migrant mortality advantage in various countries typically fail to adequately account for these features, producing mortality rates among migrants that tend to be artificially low. In such a context, the classic observation that migrants experience a mortality advantage upon arrival in the host country (often in spite of their lower socio-economic status) which diminishes over time and disappears or even reverses for the second generation, remains difficult to interpret in epidemiological terms. This challenges the ability of policy makers to adequately address the health needs of migrant populations. This research will examine the nature and magnitude of the migrant mortality advantage using three rich and unique longitudinal data sets from France, a highly-relevant context for this area of research. Specific features of these data sets will allow us to address methodological pitfalls that have plagued research on migrant mortality so far. These features include: a rigorous follow-up of deaths occurring among migrants in the host country; the possibility of censoring individuals when they leave the host country; a rigorous follow-up of deaths worldwide for former residents of the host country; and the possibility of capturing second-generation migrants. These unique features will allow us to evaluate two major hypotheses for the migrant mortality advantage: (1) the hypothesis that the migrant mortality advantage is a data artifact due to inadequate recording of deaths and person-years of exposure for migrant populations in their host country; and (2) the hypothesis that the mortality advantage is produced by selective return migration, a hypothesis also known as the salmon bias hypothesis. The focus is on mortality above age 30, with additional focus on mortality above age 65. Using data with unique features and representing a highly-relevant context, this project will evaluate hypotheses with a greater level of precision and comprehensiveness and will make important contributions to debates about how a set of general mechanisms can affect patterns of the migrant mortality advantage, with implications for a variety of contexts where a similar advantage has been observed.